Implementation Details ---------------------- This page describes how our lab implemented the procedures required by the official protocol for the RRR. It also describes and justifies any additions to or departures from that protocol. You can view the official protocl and the main project page for this RRR using these links: - Official Protocol: [https://osf.io/6wvj4/][1] - Main RRR project page: [https://osf.io/hgi2y/][2] - Project page with all materials: [https://osf.io/pkd65/][3] ---------- #### Experimenters #### The qualifications/training of those who are conducting the testing and describe how any research assistants were trained on the protocol: Our team consists of senior and junior researchers from University of Amsterdam: Prof. dr. Agneta Fischer, Director Psychology Research Institute. Expertise on facial expressions of emotions and emotional mimicry. Dr. Disa Sauter, Assistant Professor, Department of Social Psychology. Expertise on nonverbal communication of emotions. Dr. Suzanne Oosterwijk, Postdoctoral Researcher at the Department of Social Psychology. Expertise on the embodiment of emotions. Dr. Peter Lewinski, Marie Curie Research Fellow at Amsterdam School of Communication Research. Expertise on facial expressions. Lisanne Pauw, PhD Candidate at the Department of Social Psychology., Expertise on emotions Xia Fang, PhD Candidate at the Department of Social Psychology., Expertise on emotions ---------- #### Setting/Lab/Equipment #### Our departmental lab has recently moved to Roeterseiland Campus in Amsterdam. On the campus the new building REC-L is home of all labs of our faculty’s different departments. Communication Science has access to 11 cubicles, which have a computer setup to conduct tests with participants. Testing stations are arranged so that each participant cannot see or hear any other participant, and cannot see any of the testing materials for other participants. The cubicles are separate rooms with doors that can be closed, and participants cannot hear others laugh. The equipment we will use to conduct the study: *Standard PC computers *Testing cubicles *Informed consent forms in Dutch (https://osf.io/e2gjm/) *Printed information brochures. These brochures contain the cover story (i.e., the task instructions). *Practice task sheets in Dutch (https://osf.io/4efjc/). *Pens. (e.g., www.merriartist.com/Stabilo_p/sw68-46.htm www.artsupplywarehouse.com/prodDetail.php?id=18325 ). *Boxes of paper tissues *Alcohol swabs *Video camera recording system ---------- #### Sample, subjects, and randomization #### **Target sample size:** 100 **Target sample demographics:** Native Dutch-speakers, coming from undergraduate students’ participant pool of Department of Communication, University of Amsterdam. The students have thus not been taught about the facial feedback hypothesis in their studies. Around 70% female, range age: 19-23, proficient in English language, they will compensated €5 per 30min or with 0.5 credit point. **Minimum sample size after exclusions:** n = 100 **Stopping rule(s):** We will put a cap on the maximum number of participants in the computerized system to n = 70 per condition, we will test 70 people no matter what. Our ideal number of included participants is 50 per condition. If, after exclusions, we have fewer than 50 participants in either of the conditions, we will test additional sets of 10 participants using the same alternating assignment until both conditions have more than 50 participants with usable data (checking that by putting an username check from our formalized university lab pool). **Randomization to conditions:** On each brochure is a unique participant number so that participants can be assigned to conditions based on whether the participant number is odd or even. Odd numbered participants are assigned to the pout condition, and even numbered participants are assigned to the smile condition in Dutch (https://osf.io/takdv/). If we need to test additional participants, we would simply continue with the method described above until we reach requiered minimum number of participants. **Blinding to conditions:** We will follow the instructions in the original protocol. **Exclusion rules:** The official protocol describes a number of reasons for exclusion. We will be using the same exclusion rules required by the official protocol. If we exclude participants, we will retain their data, but we will mark them for exclusion. Exclusion decisions will be made by someone blind to condition assignment. We added a question to the very end of the protocol asking if participants have completed this (or similar) study before. This is because we shouldn not include participants who have previously participated. We added this additional exclusion rule because we wanted to ensure that participants who had been involved in our pre-testing or a similar experiment didn't participate again in the actual study. The questions we added are: What do you study? ____________________________________ Heb je ooit deelgenomen aan een ander onderzoek waarin je werd gevraagd om een pen in je mond te nemen en cartoons te beoordelen? JA / NEE The exclusion rules we added are: * Answered that did not study communication science
 * Participated in a study when they had to put the pen in the mouth Modified exclusion rules after 29 Oct 2015: For our replication effort we ended up with only 42 participants, which is not what we expected, in a period of 09 Sept-15 Oct. and our target was at least 140 participants no matter what as stated above. This is below our expectations. The student-assistant who ran the study was in the lab almost every hour of every day. The problem was that not enough students signed up, and that we had to reject people because they weren’t communication students. We talked to replication coordinators about it. They agreed that we can continue. To gather the requiered 140 participants no matter what, we will: a) increase the financial compensation from original 5 to 10 euro for half an hour. b) open up participation to anyone except psychology students and students who have participated before. ---------- #### Software/Code #### We will be using the provided materials, including any software or scripts, and we have verified that they work in our laboratory. ---------- #### Differences from the official protocol #### After the replication part of the experiment, we would like to add two additional extensions. Thus we will make sure that the critical replication condition comes first and that any other conditions come after the completion of the measures from the official protocol. First we would like to make a comparison between how the pen manipulation influences three different categories of cartoons. In the replication subjects will judge moderately funny cartoons. To investigate the hypothesis that embodiment effects (i.e., facial feedback) exert a stronger influence on ambiguous stimuli than non-ambiguous stimuli, we would also like to examine the effect of the pen manipulation on cartoons that are seen as very funny, or not seen as funny at all. In order to do this, we propose to add two blocks in which subjects judge high and low funny cartoons. These categories will be based on cartoon ratings taken from a pre-test. Furthermore, we would like to expand the experiment into a full within-subjects design. Thus, we would like to add three blocks in which subjects judge low funny, moderately funny and high funny cartoons, in combination with the other pen manipulation. In that way we can analyze differences between pen conditions for the three different cartoon categories fully within participants. The experiment will have 3 judgment blocks with each 4 cartoons for each pen condition (6 blocks total). We will run the following 8 different counterbalance versions of the experiment: Pen 1 – medium – high – low Pen 2 - medium – high – low Pen 1 – medium – low – high Pen 2 - medium – low – high Pen 1 – medium – high – low Pen 2 - medium – low – high Pen 1 – medium – low – high Pen 2 - medium – high – low Pen 2 – medium – high – low Pen 1 - medium – high – low Pen 2 – medium – low – high Pen 1 - medium – low – high Pen 2 – medium – high – low Pen 1 - medium – low – high Pen 2 – medium – low – high Pen 1 - medium – high – low **Note.** that the replication part of the experiment only involves the very first block of each conterbalance versions. [1]: https://osf.io/6wvj4/ [2]: https://osf.io/hgi2y/ [3]: https://osf.io/pkd65/